JPH0981770A - Projection image constituting method, artificial three-dimensional image constituting method, and projection image display device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately set the update of a viewpoint by a center projection method on a screen when an endoscopic image by the center projection method of a stack three-dimensional image is displayed. SOLUTION: The center projection of the stack three-dimensional image 3 is performed from a viewpoint E by giving the view point E and a projection surface 1. An image is obtained on the projection surface 1 as if the projection image were shaded and observed through an endoscope. Further, the stack three-dimensional image is projected on a projection surface 1 by a parallel projecting method. At this time, the projection image is cut by a center line N-N' and only the inner-side image is shaded and displayed. Thus, the images are obtained on the projection surface 2 as if the view point E and projection surface 1 were viewed from side, and the view point E is updated on this screen.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a pseudo three-dimensional image display, and more particularly to a display device suitable for dynamically obtaining an endoscopic pseudo three-dimensional image.

[0002]

2. Description of the Related Art An X-ray CT image is piled up to form a three-dimensional image, which is directly displayed on a two-dimensional screen, or the three-dimensional image is projected from a certain point on a projection surface. , There is a technology to display this projected image. It can be used directly for examinations, diagnoses, surgery, or to support them. In addition to X-ray CT images, when three-dimensional images are obtained by performing three-dimensional measurement like MRI, they are developed into two-dimensional tomographic images and used in the same field as X-ray CT images. There is a law.

[0003]

When projecting onto a projection plane, there are a parallel projection method and a central projection method. The applicant of the present patent has filed various patent applications relating to the parallel projection method, and even in the central projection method, Japanese Patent Application Nos. 6-3492 and 6-143496.
Have filed a patent application such as The central projection method is excellent in obtaining an endoscopic projection image, and its details are described in Japanese Patent Application No. 6-3492. Further, Japanese Patent Application No. 6-143496 describes the significance and method of updating the viewpoint endoscopically and an updating example.

The viewpoint update in Japanese Patent Application No. 6-143496 can be relatively easily performed by giving update coordinates in advance or setting by using a data input method such as a keyboard. However, it is not easy to display an image by the central projection method on the screen and update the viewpoint using a cursor or the like on the screen while viewing the image. For example, by the central projection method for the human bronchus,
Consider a case where an image from the front of the bronchus to the back is obtained and displayed. Since the distance on the back side is not displayed on the screen, the new viewpoint cannot be accurately set at the position on the back side. Moreover, since there is a width in the depth, it is required to decide at which position of the width the new viewpoint is set, but this is also impossible. Therefore, in order to set the new viewpoint position accurately on the display screen,
It is necessary to display an image effective for updating the viewpoint.

An object of the present invention is to obtain an image effective for updating a viewpoint by a parallel projection method, and to set and input a new viewpoint on this image. A projection image construction method and a pseudo three-dimensional image construction method. And a display device.

[0006]

DISCLOSURE OF THE INVENTION The present invention discloses a projection image construction method for obtaining a first projection image by the central projection method and a second projection image by the parallel projection method from a stacked three-dimensional image. .

Further, according to the present invention, a viewpoint and a projection surface are paired and projected onto a projection surface by a central projection method with respect to a stacked three-dimensional image, and this is shaded to obtain a pseudo three-dimensional image. A pseudo-three-dimensional image is obtained by projecting the original image onto a projection surface by a parallel projection method so that the position including the viewpoint or the vicinity of the viewpoint becomes a projection target area, and obtaining the pseudo-three-dimensional image. The following three-dimensional image construction method is disclosed.

Further, according to the present invention, the central viewpoint and the central projection plane are updated as a pair to obtain a central projection image of a three-dimensional image stacked on the central projection plane, and the projected image is shaded and displayed. In the display device, a first memory for storing a first pseudo-three-dimensional image obtained by shading the central projection image and a parallel projection plane orthogonal to the central projection plane for paralleling the stacked three-dimensional images. A second memory for storing a second pseudo-three-dimensional image obtained by extracting a pixel far from the parallel projection viewpoint plane while projecting and shading; a display;
The second pseudo-three-dimensional image of the memory is displayed on the display surface of the display, and the new viewpoint that replaces the central viewpoint of the stored image on the first memory is displayed on the second pseudo-display. A projection image display device including means for setting on a three-dimensional image and a projection image display device is disclosed.

[0009]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a diagram for explaining a central projection method and a parallel projection method using projection data in the central projection method. A plurality of tomographic images are piled up, and this is centrally projected from the viewpoint E onto the projection plane 1 (central projection plane P). This is the central projection method.
More on -3492. On the other hand, a method for obtaining a dynamic projection image by updating the viewpoint E and the projection plane 1 in correspondence with each other is described in detail in Japanese Patent Application No. 6-143496 by the present applicant.
An image obtained by the central projection method obtained in Japanese Patent Application No. 6-3492 is an image that spreads radially from the viewpoint, which is an image as seen by a so-called endoscope. That is, Japanese Patent Application No. 6-3492
Can obtain a pseudo-image as if a stacked three-dimensional image was seen with an endoscope. For example, it is convenient to see internal images of the bronchus, stomach, and intestines.

On the other hand, while Japanese Patent Application No. 6-3492 has a fixed viewpoint, Japanese Patent Application No. 6-143496 updates while the viewpoint E and the projection plane 1 are opposed to each other, and in response to this position update. We have decided to obtain dynamic endoscopic images.

On the other hand, the projection method by the parallel projection method is a method of projecting a stacked three-dimensional image in parallel. It is convenient for obtaining the appearance image of the stacked three-dimensional image and the cut image when cut at the cut surface.

Even in the central projection method and the parallel projection method, the pixel value obtained on the projection surface is not the pixel value itself on the stacked three-dimensional image, but a small pixel value (low luminance value) if the distance is long, or a short pixel value if the distance is short. It is an image that is shaded such that it has a large pixel value (high brightness value) (display is such that distant objects appear dark and nearby objects appear bright). For this shading, various methods such as the Debs method, the surface method, and the volume rendering method are adopted. Furthermore, multiple pixels may be projected to the same projected pixel position (1
One projection point has projections from multiple projection target points). This is so-called pixel overlap. Such overlap occurs due to the presence of two or more projection target pixels in the depth direction at the same projection pixel position. Therefore, the projection target pixel on the back side is deleted, and only the projection target pixel on the front side (or the back side) is selected as the projection pixel at this position. This is called concealment processing. As a result, when they overlap, the projection target pixel on the front side is projected and displayed.

The projected image obtained by such shading and concealment is not the pixel value itself of the original stacked three-dimensional image but the pseudo pixel value, and when the projected image is displayed on the display surface. It is called a pseudo-three-dimensional image because it is apparently three-dimensional.

In FIG. 1, the projection plane 2 is a parallel projection plane Q by the parallel projection method. The projection target image at this time is a stacked three-dimensional image used for projection by the central projection method. The projection plane 2 may be in any spatial position, but the one having the highest utility value is the projection plane in an angle direction parallel to the orthogonal projection line that is perpendicularly incident from the viewpoint E to the projection plane 1, that is, The projection planes 1 and 2 are orthogonal to each other. At this time, an endoscopic image viewed from the viewpoint E is obtained on the projection surface 1, and a parallel projection image viewed from a direction orthogonal to the orthogonal projection line from the viewpoint E is obtained on the projection surface 2. .

This positional relationship is shown in FIG. FIG. 2 is an image of a living tissue a ₁ having a cavity inside, and the living tissue a ₂ has the cavity branched into two (b ₁ and b ₂ ). Such an image is, for example, a bronchus. The viewpoint E is placed in the cavity and the projection plane 1
The central projection is performed on and a pseudo-three-dimensional image is obtained. On the other hand, any position in the direction parallel to the orthogonal projection line (that is, parallel position)
The projection plane 2 is set to, and the pixel data of the pixel position (that is, the projection target point) used in the projection onto the projection plane 1 is projected onto the projection plane 2.

As a result, an endoscopic image viewed from the viewpoint E is obtained on the projection surface 1, and a stacked three-dimensional image is laterally projected on the projection surface 2 at a position parallel to the orthogonal projection line. A parallel projection image when viewed can be obtained. Of course, there can be any angle example other than parallel to the orthogonal projection line.

Further, the projection planes 1 and 2 may be considered as the projection memories themselves corresponding to each of them, and the projection point can be defined by the address of the projection memory. Further, when the projection memory is larger than the image memory, an image of the image memory size is cut out (read out) from the projection memory, stored in the image memory, and displayed. If the projection memory and the image memory have the same size, the contents of the projection memory are displayed as the image memory as they are or are transferred to the image memory as they are and displayed.

A method of using the parallel projection image in the present invention will be described. First, the position of the endoscopic image viewed by the central projection method from the lateral direction (that direction when viewed from the front side to the back side of the paper in FIG. 2) or It is used to visually observe what kind of situation it is. Secondly, it is used to specify the position of the viewpoint when updating (forward or backward) the viewpoint in Japanese Patent Application No. 6-143496. And so on.

The second method is to use it for the viewpoint update setting, since it is possible to know in which direction the viewpoint should be updated when viewed from a parallel projection image from the horizontal direction. Is. In particular, the viewpoint is rarely updated on a straight line at regular intervals, and for example, as shown in FIG. 3A, a curvilinear update locus is taken as E ₁ → E ₂ → E ₃ → E ₄ → E _5. There is also. In such a case, just displaying the central projection image cannot grasp the movement in the depth (or front) direction, and it is difficult to update the viewpoint as shown in FIG. It is necessary to rely on the feeling or to obtain the positions of E ₁ , E ₂ ′ ... E _{5 in} advance and update according to these positions.

The loci of E _{1 to} E ₅ may be performed on one plane, but may be on different planes.
FIG. 3B shows such an example, and E ₂ ′ and E ₄ ″ are updated to different planes. That is, E ₁ , E ₃ and E ₅ are shown in FIG.
It is each point on the same plane as (a), and E ₂ ′ and E ₄ ″ are examples of trajectory updates to other planes different from that plane. On a projection plane 2 parallel to the orthogonal projection line as shown in FIG. If so, Fig. 3 (a)
Is suitable for updating such that the updating plane moves parallel to the plane. Therefore, in an example in which another plane parallel to the projection plane 2 is used as the update plane, the projection plane 2 is preferably selected so as to be parallel to the update plane (therefore, in this case, the orthogonal projection line). Is not always parallel). In addition, in the example shown in FIG.
Is preferably selected as an average surface (inclination) between the surfaces formed by E ₁ , E ₂ and E ₃ and the surfaces formed by E ₂ ′ and E ₄ ″.

To set the update viewpoint from the horizontal direction, a parallel projection image viewed from the horizontal direction is displayed on the screen, the update position is set on the image, this is read by the mouse cursor, and latched in the memory. Let's keep it. Details will be described later in the description of FIGS. 5 and 6.

FIG. 4 shows an endoscopic display image obtained by the central projection method. FIG. 5 is a diagram showing the relationship between the viewpoint E ₁ and the three-dimensional original image 11 and the projection plane 12 when the image of FIG. 4 is obtained, and FIG.

FIG. 5 shows a plurality of CT slice planes (..., # _i ,
...) to obtain a three-dimensional original image 11 stacked in parallel, this giving viewpoint E ₁ and the projection plane 12, the viewpoint E ₁ on the projection plane 12
It is supposed that an endoscopic image is obtained by the central projection method from. The three-dimensional original image 11 includes an internal cavity portion C separated by inner walls a and b and external tissues d ₁ and d ₂ (hatched notation).
Is an example of an image with and the hollow part C is 2
It branches into _two cavities C ₁ and C ₂ .

The projected image on the projection plane 12 from the viewpoint E ₁ is written in the memory M ₁ (described later), and its display image is shown in FIG.
Becomes Here, in FIG. 5, a hatched portion on the original image from the viewpoint E toward the projection surface 12 is a projection target portion in the central projection method. However, in the actual projection, since the hidden surface processing is performed, the parts A ₁ , A ₂ and A _{3 that} are the frontmost from the viewpoint E ₁ in the shaded part form the actual projected image. To do. A ₁ , A ₂ and A _{3 in} FIGS. 4 and 5 indicate the same portion.

FIG. 6 shows the parallel projection image. This is an image written in the memory M ₂ (described later). It is the figure which looked at FIG. 5 from the vertical direction (lateral direction) to the paper surface. In FIG. 6, the thick line portion and the shaded portion are not shown in the drawing, and the applicant has filled in for correspondence of FIG. In the image of FIG. 6, the right-hand direction is the depth and the left-hand direction is the vicinity of the current viewpoint E ₁ . Therefore, in order to set a new viewpoint in the depth direction, see FIG.
Of E indicated by the mouse on the _second as the screen and displays the X mark, which reads the new viewpoint E ₂ coordinates by specifying the mouse. A new projection plane corresponding to the read new viewpoint E ₂ is set, and a central projection image from the new viewpoint E ₂ is formed on the new projection surface.

As described above, on the screen of FIG. 6, an image in which the updating direction of the viewpoint is viewed from the lateral direction is displayed, and the new viewpoint can be set accurately on this image.
In addition, in FIG. 6, the new viewpoint is one point of E ₂ , but there are two points or three points.
Points or the like may be set on one image shown in FIG. Of course, an image as shown in FIG. 6 may be obtained for a new projection image obtained based on the new viewpoint E _2, and a new viewpoint E ₃ may be set from there. Hereinafter, the present invention will be specifically described.

Now, in FIG. 1, the coordinate system of a stacked three-dimensional image is given by x, y, z. Here, y is a coordinate in the stacking direction, and x and z are pixel coordinates on each CT image plane.
The coordinate system of the central projection plane is given by X and Y. Furthermore, the coordinate system of the parallel projection plane is given by η and ξ. Each coordinate and its coordinate position shown in FIG. 1 are defined below. The coordinates of the viewpoint E are (x ₁ , y ₁ , z ₁ ), and the orthogonal point C ₁ from the viewpoint E to the central projection plane is the coordinates (x _c1 , y _c1 , z _c1 ) in the x, y, z coordinate system. ), In the X, Y coordinate system (X _c , Y _c ),
Projection target point S of the CT image on an arbitrary center projection line L
Is (x ₀ , y ₀ , z ₀ ), and the projection point R on the central projection plane on the central projection line L is (x, y, z in the x, y, z coordinate system.
z), in the X, Y coordinate system (X, Y).

The other relevant parameters in FIG. 1 are as follows. a, b, c ... A parameter indicating the positional relationship between the central projection plane P and the x, y, z coordinate system, where a is the intersection of the x axis of the x, y, z coordinate system and the central projection plane P. Is the distance from the origin to the intersection, b is the distance from the origin to the intersection on the y-axis, and c is the distance from the origin to the intersection on the z-axis. γ ₁ , γ ₂ ... γ ₁ are perpendicular to the parallel projection plane Q from the origin of the x, y, z coordinate system and the xz plane (CT
The angle with the fault plane). γ ₂ is an angle formed by the straight line obtained by further projecting the perpendicular line on the xz plane and the x axis. α, β ... α are the angles formed by the projected straight line when the perpendicular line drawn from the origin of the x, y, z coordinate system to the central projection plane is projected on the zx plane. β is the angle that the perpendicular forms with the xz plane.

Further, the central projection plane P actually corresponds to a projection memory, which is referred to as a memory M ₁ . Further, the parallel projection plane Q actually corresponds to a projection memory, which is referred to as a memory M ₂ . When the scale of the memory M ₂ is small, the scale of the coordinates is scaled up to the scaled-up memory M ₂ ′.

FIG. 7 shows a memory M ₂ for storing an image on the parallel projection plane Q. The horizontal axis is η and the vertical axis is ξ. Memory M ₂ is 4
It consists of one memory 40, 41, 42, 43,
The memory 40 stores the pixel value DP ₁ at each pixel coordinate under the η-ξ coordinate system, and the memories 41, 42, and 43 store the CT image when the pixel value DP ₁ is a projection point. The x, y, z coordinate values x ₀ , y ₀ , z ₀ at the projection target point on 3 are stored.

FIG. 8 shows a memory M ₃ for storing an image on the parallel projection plane Q. It consists of four memories 44, 45, 46, 47. The memory M ₂ stores the parallel projection pixel value DP ₂ from the image position farthest from the viewpoint plane and the x, y, z coordinate values x ₀ , y ₀ , z ₀ at the projection target point in parallel projection. Store. Specifically, the parallel projection pixel value DP ₂ from the image position closest to the viewpoint plane in parallel projection is stored in the memory 44 and the x, y, and z coordinate values x ₀ at the projection target point.
The y ₀ and z ₀ are stored in the memories 45, 46 and 47.

Here, the definitions of the farthest and the closest are described. Looking at one parallel projection line, the projection line penetrates a plurality of CT slice planes, and a plurality of intersections appear.
These plural intersections are points to be projected on the parallel projection plane. Among the plurality of intersections, the closest intersection (projection target point) from the viewpoint plane (that is, the parallel projection viewpoint plane)
Is defined as the closest pixel position. On the contrary, the farthest intersection point (projection target point) from the viewpoint plane is defined as the farthest pixel position. The same definition is applied to all parallel projection lines. Then, the parallel projection pixel value DP ₁ when the farthest image position is set as the projection target point is stored in the memory 40 of the memory M ₂ .
The memories 41, 42, and 43 have x, y, and
The z coordinate values x ₀ , y ₀ and z ₀ are stored. Further, the projection pixel value DP ₂ when the closest pixel position is set as the projection target point is stored in the memory 44 of the memory M ₃ , and the memories 45, 46, 47 are stored.
The x, y, z coordinate values x ₀ , y ₀ , z ₀ of the projection target point.
To store.

The significance of the above storage in the memories M ₂ and M ₃ will be described with reference to FIG. FIG. 9A shows a cylindrical region image 20, and N-N 'is a left end portion of the image 20 when a parallel projection line is set from the front right direction to the left direction of the paper. It is the image center line of the parallel projection line passing through. FIG. 9D shows a detailed example thereof. When a parallel projection line as shown in the figure exists parallel to the image 23 (parallel to the paper surface), the image 23
And intersect at points 1, 2, 3, and 4. Then, the center positions of the points 1 and 2 and the center positions of the points 3 and 4 become the image center points. An image center line N-N 'is formed by connecting the image center points.

FIG. 9B shows the image 21 on the far side from the image centerline N-N ', and FIG. 9C shows the image centerline N-N'.
An image 22 in the foreground is shown. In FIG. 9D, NN ′
The extracted image only on the left side (back side) of the image is the image 21, and the extracted image only on the right side (front side) of NN ′ is the image 22.

The image shown in FIG. 9B is the image stored in the memory M ₂ , and the image shown in FIG. 9C is the image stored in the memory M ₃ . That is, the image at the far position stored in the memory M ₂ shown in FIG. 7 is a cut image on the far side when the image is cut at the image center line when the image is viewed from the line-of-sight direction. This corresponds to the image 21 shown in 9 (b). On the other hand, the image stored in the memory M _{3 of} FIG. 8 is a cut image on the near side when the image is cut at the image center line, and this corresponds to the image 22 shown in FIG. 9C.

FIG. 10 shows an example of a processing flow chart of the present invention. First, the flow F ₁ clears the memory M _{1 for} the central projection plane and the memory M ₂ for the parallel projection plane to zero. Flow F ₂
Is the maximum value (eg 999) in the memory M ₃ of the parallel projection plane.
9) Initial set. The maximum value is set
This is because when the cutting line is the center line of the image, the image before (close to) the cutting line is stored in the memory M ₃ .

The flow F ₃ sets initial coordinates X and Y of the display surface on the central projection surface (that is, the memory M ₁ ). FIG. 11 shows an example of cutting the display surface. Orthogonal point C from viewpoint E
FIG. 11 shows an example in which ₁ is the center of the screen and (vertical image size number) × (horizontal image size number) = m × n. The center of the orthogonal projection point C ₁ is the center E of the viewpoint E.
This is because they are selected so as to be orthogonal to each other, and the orthogonal position is the orthogonal point C ₁ . The initial coordinates X and Y are m × n
According to the following formula, which is the upper left end position in the size cut image.

[Equation 1] Here, X _c1 and Y _c1 are the coordinates of the orthogonal point C ₁ .

The flow F ₄ specifies the CT image closest to the viewpoint E. The CT image closest to the viewpoint E is the tomographic plane # (i-2) in the example of FIG. In the flow F ₅ , the intersections (that is, projection target points) x ₀ and z ₀ on the CT image when the central projection lines are radially formed with respect to the closest CT image are calculated. In terms of the relationship with the flow F ₃ , when X and Y of (Equation 1) are projection points, the projection target points x ₀ and z ₀ are obtained. The coordinate conversion formulas from x ₀ , z ₀ to X and Y in (Equation 1) will be described later.

The CT closest to the viewpoint E in the flow F ₄
The image is designated by comparing the y-coordinate value of the CT slice plane (that is, the CT image) with the y-coordinate value of the viewpoint E, and the CT slice that is closer to the projection plane than the viewpoint and is closest to the y-coordinate value of the viewpoint E. Face y
It is specified by finding the coordinate value.

The flow F ₆ checks whether or not the CT pixel values of the projection target points x ₀ and z ₀ obtained in the flow F ₅ satisfy the threshold value. Here, the threshold value is a value serving as an extraction criterion, and if the threshold value condition is satisfied, the process moves to flow F ₉ , and if not, the process moves to flow F ₇ . The threshold is
For example, when it is desired to extract a region of interest in an organ, the value is such that the region of interest can be extracted.

In the flow F ₇ , the next closest tomographic image (tomographic plane) (# (i-1) in the example of FIG. 5) is designated. Then, all the images (in the example of FIG. 5, # (i-2) to # (i +
Or specify all images) leading to 7) (i.e., the flow of the last image # (i + 7) whether finished specification of) F ₈
If not, flow F ₅ is entered. If the last image # (i + 7) has been designated, flow F ₁₆
Move on to.

In the flow F ₉ , the distance D between the viewpoint E and the projection target point S is obtained. The distance D is obtained from the following equation.

[Equation 2]

In the flow F ₃₀ , the projection position R of the memory M ₁
The distance D is written in (X, Y). At this time, when projection is performed from a plurality of projection target points to one projection position, a plurality of D is obtained from the plurality of projection target points, but the minimum distance D among the plurality of distances D is the final Writing (this is a process for removing a projection target point at a long distance, and is a hidden surface process).

In the flow F ₁₀ , the projection target point S (x ₀ ,
y _0, z ₀₎ parallel projection point of the projection plane parallel on the Q from (eta _0, the xi] _0), obtained by (Equation 3).

[0045]

(Equation 3) Here, γ ₁ and γ ₂ are the values defined above. Furthermore, when the scale of the parallel projection plane Q is changed (enlarged or reduced), the scale conversion suitable for that is performed. The parallel projection plane after the scale conversion becomes the memory M ₂ ′. An example in which scale conversion is not performed will be described below, but both can be adopted. For example, when the projection target such as inside a blood vessel is small, scale expansion is performed.

In the flow F ₁₁ , the projection target point (x ₀ , y ₀ ,
The distance DP between z ₀ ) and its parallel projection point (η ₀ , ξ ₀ ).
Is calculated by the following equation.

(Equation 4) Here, D ₀ is the CT tomographic plane to which the projection target point S belongs and y
It is the distance between the intersection with the axis and the parallel projection plane Q. y axis is C
Since it is set at the center position of the T slice plane, the intersection with the y axis indicates the center position of the CT slice plane.

[0047] Flow in F _12, the position of the memory M ₂ (η _0,
The magnitude of the pixel value of ξ ₀ (which is the pixel value DP ₁ of the memory 40 in FIG. 7) and the DP of (Equation 4) are compared. If the memory M ₂ remains zero-cleared in the flow F ₁ ,
Since the comparison result is always DP> 0, the flow moves to the flow F ₁₄ and the pixel value D is set at the position (η ₀ , ξ ₀ ) of the memory M _2.
P is overwritten (overwriting means erasing a past DP and writing a new DP). In general, each time a new DP is obtained, it is compared in size with the previous DP, and if the new DP is large, this new DP will take over. By providing this flow F ₁₂ , the largest DP is finally stored for each projection point in each pixel of the memory M ₂ . Here, the largest value means that, when parallel projection is performed on one projection point from a plurality of projection targets, in DP for a plurality of projection target points,
Refers to the maximum DP. Further, as a result of the comparison, if the new DP is smaller than the past DP, the value of the past DP is left as it is on the memory M ₂ .

In the flow F ₁₃ , the projection position R of the memory M ₃
(X, Y) pixel value (pixel value DP ₂ of the memory 44 of FIG. 8)
And the DP of (Equation 4) are compared. If DP is small, the flow moves to F ₁₅ and the projection position R (X, Y) of the memory M ₃ is overwritten with this DP. On the other hand, if the DP is large, it is not overwritten and the past DP is left as it is.
Since the memory M ₃ is initially set to the maximum value in the flow F ₂ , the new DP always replaces the maximum value in the first comparison after the initialization, and the smaller DP is set in the subsequent comparison. Will be overwritten. By providing this flow F ₁₃ , the smallest D in the memory M _3.
P can be stored for each projection point.

In the flow F ₁₆ , the coordinate X on the projection plane P (memory M ₁ ) is updated, and the flow F ₄ to the flow F ₁₅ are repeated until the maximum coordinate value (upper right position of the cutout size + 1) is reached. When the maximum coordinate value is reached, the flow F ₁₈ updates the coordinate Y and returns the coordinate X to the original coordinate (the left end X coordinate of the cutout size). Thereafter, similar processing is repeated until Y reaches the maximum coordinate value (Y coordinate of the lower end portion of the cutout size) (Flow F ₁₈ , F ₁₉ ).

In the flow F ₂₀ , instead of all the pixel values D of the memory M ₁ obtained in the flow F ₃₀ , the pixel value D ′ of the following equation is written. Dmax is the maximum pixel value (setting value).

(Equation 5) This pixel value D'is a pixel value that is shaded by the depth method. Other shadings are possible.

Further, in the flow F ₂₀ , the flows F ₁₂ and F _{14 are}
The pixel value DP ′ of the following equation is written in place of all the pixel values DP of the memory M ₂ obtained in step a. DPmax is the maximum pixel value (setting value). This pixel value DP 'is a pixel value that is shaded by the depth method. Other shading can be used.

(Equation 6) Thus, the memory M ₁ obtained hidden surface of is and shadowing image, in the memory M ₂ farthest viewed from the cut surface,
Moreover, a shaded image is obtained.

The image in the memory M _{1 and} the image in the memory M ₂ obtained in step F ₂₀ of FIG. 10 are sent to the display device and displayed as images. Here, the image in the memory M ₁ becomes an image as observed by an endoscope, and the image in the memory M ₂ becomes an image as shown in FIG. The position of the current viewpoint E ₁ is displayed on the image of FIG. 6, and an arbitrary position different from that is referred to as the new viewpoint E ₂ with reference to this viewpoint E ₁ . The computer can capture the new viewpoint E ₂ by pointing it with the cursor, and its position (η _i , ξ _i ).

FIG. 12 is a processing flowchart for setting the new viewpoint E ₂ , reading the coordinates, and converting the coordinates. In the flow F ₂₁ , one point on the screen of FIG. 6 is designated by the cursor as the new viewpoint E ₂ . In the flow F ₂₂ , the designated coordinates (coordinates on the parallel projection plane) (η _i , ξ _i ) are read. If scale conversion is necessary, the read coordinates are scaled (scaled) to the coordinates before scale conversion.

Further, in the flow F ₂₃ , designated coordinates (η _i ,
ξ _i ) from the memories 41, 42, 43 of the memory M ₂ of FIG.
The corresponding coordinates (x ₀₁ , y ₀₁ , z ₀₁ ) above are read out. This is because the designated and read coordinates (η _i , ξ _i ) are the vertical and horizontal addresses of the memory 40 of the memory M _{2 in} FIG. 7, and the memories 41, 42 and 43 have CTs corresponding to the addresses.
Points (projection target points) on the image are stored, and the corresponding projection target points (x ₀₁ , y ₀₁ , z) are accessed by accessing the memories 41, 42, and 43 with coordinates (η _i , ξ _i ). _This is because ₀₁ ) can be read.

In the flow F ₂₄ , the corresponding coordinates (x ₀₂ , y ₀₂ , z ₀₂ ) on the memories 45, 46, 47 of the memory M _{3 in} FIG. 8 are read from the designated coordinates (η _i , ξ _i ). Memory 4
5, 46, 47 are addressed by the same address (η, ξ) as the address (η, ξ) of the memory 44, and the projection target point () corresponding to the address (η, ξ) is stored in the memories 45, 46, 47. Since x ₀₂ , y ₀₂ , and z ₀₂ ) are stored, they can be read.

In the flow F ₂₅ , the average value of the two read projection target points (x ₀₁ , y ₀₁ , z ₀₁ ) and (x ₀₂ , y ₀₂ , z ₀₂ ) is taken and regarded as the center point. (X ₀₁ , y ₀₁ , z
₀₁ ) is the farthest point, and ( _x02 , _y02 , _z02 ) is the closest point, and the average value thereof is the central point of both. Such processing of the flow F ₂₅ is useful in the following examples. It is assumed that the blood vessel or the trachea has a predetermined inner diameter, the farthest point with respect to the inner diameter is stored in the memories 41, 42, 43, and the closest point with respect to the inner diameter is stored in the memories 45, 46, 47. And focusing on the new viewpoint E ₂ that has been set,
There is a predetermined cavity in the depth direction from the paper surface of FIG. 6, and the center point of the cavity is optimal as the new viewpoint E ₂ ′. The process for obtaining this center point is flow F ₂₅ .

The display of the memory 44 of the memory M ₃ is not always necessary. This memory 44 is just a flow F ₂₅
This is because it is used as a temporary buffer memory for finding the center point in. However, when the image in front of the cut surface (FIG. 9C) is desired, it may be displayed on the display screen.

Pixel position S in the flow F ₅ of FIG.
The relationship between the projection target point S and the projection point R for projecting (x ₀ , y ₀ , z ₀ ) onto the projection point R (x, y, z) and R (x,
A conversion formula for converting y, z) into coordinates R (X, Y) of the coordinate system on the projection plane will be described below.

First, when the projection plane P is defined by the x, y, z coordinate system,

(Equation 7) It is represented by Also, point E (x ₁ , y ₁ , z ₁ ) and point P (x,
A straight line 22 passing through (y, z) is in the x, y, z coordinate system,

(Equation 8) Given in.

Since the projection plane P is orthogonal to the viewpoint at C ₁ point (x _c1 , y _c1 , z _c1 ),

[Equation 9] The relation between x, y, z of R (x, y, z) and X, Y when the coordinates of this point R on the projection plane X, Y coordinate system are R (X, Y) Becomes

(Equation 10) Here, at the point C ₁ (x _c1 , y _c1 , z _c1 ), for example, a point where the perpendicular line drawn from the viewpoint E (x ₁ , y ₁ , z ₁ ) to the projection plane P intersects the projection plane P ( The distance between this point and the viewpoint E is h),

[Equation 11] May be used.

Although the relationship between the projection plane and the display screen has been described above, the projection plane is a projection memory defined by the coordinate system XY, and the whole or a part of this projection memory is displayed in the display memory. And display it on the display screen.

When the image projected on the projection plane P is displayed on the display screen (not shown) of the display device with 512 pixels × 512 pixels in the vertical direction, X and Y take values from −256 to +256. For each of X and Y, x, y, and z are determined by (Equation 10). Since x ₁ , y ₁ , and z ₁ of the E point are arbitrarily given, the coordinates x ₀ and z ₀ of the pixel S point are determined on the tomographic image of y = y ₀ by (Equation 12).

(Equation 12)

The coordinate conversion as described above is performed for all points on the projection plane P corresponding to the display screen. In addition, it is performed for all tomographic images 3.

FIG. 13 shows a system diagram of the projection display device of the present invention. This projection display device includes a CPU 50, a main memory 51, a magnetic disk 52, a display memory 53, a CRT.
54, controller 55, mouse 56, keyboard 57
And a common bus 58. Each tomographic image is stored in the magnetic disk 52, and the CPU 50 performs predetermined processing according to the projection display software (FIGS. 10 and 12) in the main memory 51. In this processing, input / output processing and processing operations are performed using the mouse 56 and the keyboard 57. The stacked three-dimensional image is displayed on the CRT 54 via the display memory 53, and is operated by an operator or the like as shown in FIGS.
Process 2 is performed. Then, the contents of the memory M ₂ or M ₂ ′, which is a part of the main memory 51 or the magnetic disk 52, are displayed, and the operator views the contents and updates the viewpoint. The endoscopic images newly obtained in the memory M ₁ (which is also a part of the main memory 51 and the magnetic disk 52) by updating the viewpoint appear on the screen one after another to provide the treatment plan information. It is also possible to provide the operator with various information such as diagnosis, examination, and treatment. Further, the display contents are stored in the magnetic disk 52 and used for re-display.

Although a CT tomographic image is used, it can be applied to an MRI image. Although the parallel projection image has been described on the viewpoint updating side, there is an example in which it is displayed as a mere observation target image.

[0066]

According to the present invention, the original image data used when the image is projected from the viewpoint to the central projection plane and shaded,
An image with depth information obtained by parallel projection can be obtained.

[Brief description of drawings]

FIG. 1 is a diagram illustrating the principle of the present invention.

FIG. 2 is a diagram illustrating an example of using the central projection method and the parallel projection method of the present invention.

FIG. 3 is a view update example diagram.

FIG. 4 is a diagram showing an image display example obtained by the central projection method.

FIG. 5 is an explanatory diagram of a central projection method.

FIG. 6 is a parallel projection view of the image of FIG.

FIG. 7 is a diagram showing the contents of a memory M ₂ .

FIG. 8 is a diagram showing the contents of a memory M ₃ .

FIG. 9 is an explanatory diagram of a cutting process by NN ′ (center line).

FIG. 10 is a processing flowchart diagram according to the central projection method and the parallel projection method of the present invention.

FIG. 11 is a correspondence diagram between the central projection plane and the display screen.

FIG. 12 is a flowchart for calculating coordinates of a new viewpoint on an image obtained by the parallel projection method.

FIG. 13 is a processing system diagram of the present invention.

[Explanation of symbols]

1 center projection plane Q 2 parallel projection plane P 3 CT tomographic plane M ₁ center projection image memory M ₂ parallel projection image memory M ₂ ′ parallel projection image memory after scale conversion E viewpoint 50 CPU 51 main memory 52 magnetic disk 53 display memory 54 CRT 55 controller 56 mouse 57 keyboard

Claims (3)

[Claims] 1. A projection image constructing method for obtaining a first projection image by a central projection method and a second projection image by a parallel projection method from stacked three-dimensional images. 2. A stacked three-dimensional image is projected onto a projection surface by a central projection method with a viewpoint and a projection surface as a pair, and this is shaded to obtain a pseudo-three-dimensional image. On the other hand, a pseudo-three-dimensional image is obtained by projecting onto the projection surface by the parallel projection method so that the position including the above-mentioned viewpoint or the vicinity of the viewpoint becomes the projection target area, and shadowing this to obtain a pseudo-three-dimensional image. Image construction method. 3. A projection image display device adapted to update a central viewpoint and a central projection plane in pairs to obtain a central projection image of a three-dimensional image stacked on the central projection plane, and display this by shading it. , A first memory for storing a first pseudo-three-dimensional image obtained by shading the central projection image, and a parallel projection plane orthogonal to the central projection plane for projecting the stacked three-dimensional images in parallel. A second memory for storing a second pseudo-three-dimensional image obtained by extracting pixels farther from the parallel projection viewpoint, a display, and a second pseudo-three-dimensional image of the second memory. A means for displaying an image on a display surface of a display device and setting a new viewpoint on the displayed second pseudo-three-dimensional image instead of the central viewpoint of the stored image on the first memory; A projected image display device.